Identificación y Control Wavenet de un Motor de CA

Abstract

This paper presents a control scheme to tune online the proportional, integral and derivative gains of a discrete PID controller, through the identification and control of a SISO stable and minimum phase dynamic system. This is accomplished using a radial basis network neural with daughter Morlet wavelets activation functions in cascaded with an infinite impulse response (IIR) filter. This scheme is applied in real time to control the speed of an AC three-phase induction motor supplied with a three-phase inverter. So in this way we show how the identification and control scheme can be implemented in this type of plants that are widely used in industry, without the need of mathematical model parameters of the induction motor. We present numerical simulation and experimental results.